fix: add 'mode' back to 'ai_ask' schema (#435)

openapi.json

@@ -14,7 +14,7 @@
       "url": "http://www.apache.org/licenses/LICENSE-2.0"
     },
     "version": "2.0.0",
-    "x-box-commit-hash": "b9f1a9f298"
+    "x-box-commit-hash": "0daa9bfee9"
   },
   "servers": [
     {
@@ -24023,9 +24023,20 @@
         "x-box-tag": "ai",
         "required": [
           "prompt",
-          "items"
+          "items",
+          "mode"
         ],
         "properties": {
+          "mode": {
+            "type": "string",
+            "description": "The mode specifies if this request is for a single or multiple items. If you select `single_item_qa` the `items` array can have one element only. Selecting `multiple_item_qa` allows you to provide up to 25 items.",
+            "enum": [
+              "multiple_item_qa",
+              "single_item_qa"
+            ],
+            "example": "multiple_item_qa",
+            "nullable": false
+          },
           "prompt": {
             "type": "string",
             "description": "The prompt provided by the client to be answered by the LLM. The prompt's length is limited to 10000 characters.",